Here's the falsifiable bit: expect Trumps poll numbers among evangelicals to improve in the next few weeks, and up to the election assuming he doesn't say anything stupid about Jesus. Specifically, look for the number of evangelicals who strongly disapprove of Trump to start dropping.
My anecdotal evidence and gut feeling will be incorrect if the number of evangelicals who strongly disapprove of Trump to remain the same or increase (again, assuming that Trump doesn't say his daughter is the second coming of Christ or something equally offensive).
You can also call me an idiot if Trump says his daughter is the second coming of Christ and still improves his favorability among evangelicals.- Killing Donald Trump Part Four: Really, Guys? Really? (March 14, 2016)
Now obviously, I was correct about Trump improving his numbers with Evangelicals. He beat out Bush II's percentage for his first term among all Protestants (58% to 56%) and took fully 80% of the White Evangelical vote (couldn't find numbers for various non-white evangelical voters). The number of Evangelicals who supported Trump also went up significantly from April to November (only 15% of regular evangelical churchgoers supported Trump in April to 46% Strongly Support in November).
So the result was correct. What was wrong? The model, the "why this result will happen" was ambiguous. I expected Trump to gain evangelical support because of the riots. There was no accounting for other issues that might shore up his support.
Trump would likely have gained an Evangelical boost towards the end regardless since Hilary Clinton took the Dem nomination, Voting for Hilary Clinton would be even more of a betrayal to Evangelicals' self-image than voting for Trump. My model didn't account for that.
Even when you're right, the goal isn't to be right. The goal is to make your model of reality closer to reality. Scott Adams is right about Predicting rather than Rationalizing. If your model predicted incorrectly, or you were right for reasons outside of your model, you're wrong.
I'm going to try to include the why more into my predictions in the future. This time, I went off a gut feeling,
That said, I have one more gut feeling I need to pass from my stomach before moving onwards and upwards. It's about White Evangelicals, and we'll look at it after the break.
I have a gut feeling that the 20% of White Evangelicals who voted against Trump were the people in the pulpits, not the people in the pews. In other words, 20% of White Evangelicals that voted against Trump were largely pastors, elders, deacons, and whatever other leadership roles exist in their denominations. This tendency would be most strong among "Celebrity Christians."
The rationalization for this gut feeling comes down to two factors: money and education.
Non-pastoral leaders (deacons, elders) are social enough to be elected and wealthy enough to have free time and live "Christian lifestyles." This means they likely work in professions less effected by the stagnate economy. This means they're likely educated. This means their social group includes more people who think Trump is literally Hitler. If they want to continue in these social groups, they must denounce Trump.
This gets into the social dimension of Distributed Thoughtware: the people you associate with literally block you off from certain forms of thinking.
Pastors don't generally make much money, but they do tend to go to seminary and they also have to deal with non-pastoral leaders. Basically, they're educated people who rub elbows with educated people, who are rubbing elbows with a lot of people who think Trump is literally Hitler.
The people in the pews, on the other hand, are by definition less capable/interested in being elected. Maybe they don't have much free time or feel unqualified by lack of education. This means they're more likely to be poor and/or uneducated. This means they're more likely to work in industries affected heavily by the stagnate economy (low real wages). This means they're surrounded by the people most likely to benefit from Trump's policies.
But that's all about the past election. What about the future prediction? Well, that's a gut feeling based on a gut feeling, and we're going to look at it in a future post.